2026 ASEE Annual Conference & Exposition

Students need feedback about their performance throughout any course, but particularly in capstone design because (1) a large team project is different from lectures and labs, (2) a failure in a two-course sequence is more costly, and (3) the skillset for success is different from other courses. A significant portion of a typical capstone course grade is normally evaluated at or near the end of the term, whether in the form of a final report, completed prototype, or summative design presentation. This paper evaluates how well certain grading metrics – for which students do receive feedback throughout the course – predict a student’s final grade.

The metrics used to evaluate final course grades in a Capstone Design project at York College of Pennsylvania are: (1) brief periodic stand-up presentations, (2) “milestone” tasks completed according to an instructor-defined timeline, (3) written summative technical report, (4) poster created within project sub-teams, and (5) professionalism evaluated by both instructor and peers. Students receive feedback throughout the courses from the first and second metrics listed. The third and fourth are end-of-semester deliverables. The last is typically evaluated at the end of the semester as well. Data is available and analyzed across eight years with a consistent project (Formula SAE), grading system, and instructor. There are typically about 20 students, distributed across sub-teams of three to five members, leading to a total of 164 individuals in the data set.

Results show that the first two metrics above, which provide students feedback throughout the course, are reasonably good tools for predicting final grades: they are ranked second and third out of the five measures used. In particular, periodic standup presentation grades (metric 1) predict the final course grade within one half letter grade 84% of the time and “milestone” tasks (metric 2) do so 85% of the time. This is better than the correlation of two end-of semester deliverables: metric 3 at 70% and metric 4 at 69%. The best indicator, however, of final grade status is professionalism (metric 5) at 93%. Professionalism has typically not been evaluated until the end of the semester. Fortunately, it is easy to conduct some professor and peer evaluation earlier in the term to provide students that valuable feedback. In addition to the method just described – prediction within one half letter grade – this paper uses several techniques to measure the correlation of grading metrics against final course grades and evaluate whether students are receiving adequate feedback throughout the semester.